Chapter 3
Total lipid extracts from characteristic soil horizons in a podzol
profile
The podzolization process is studied through lipids in 9 characteristic podzol
horizons. Organic matter accumulates particularly with aluminium in the Bh horizon,
while the hard, cemented Bs horizon below this is mainly formed by iron oxides. The
low soil pH seems to have no great influence on the preservation of lipids as reflected
by the absolute amounts present (mg·g-1 soil) and the presence of bacterial lipid
markers throughout the profile. Independent of soil pH, lipids accumulate in
organically enriched horizons. Albeit, high molecular weight organic compounds
accumulate to a relatively greater extend than lipids in these horizons. A lipid signal
related to the aerial parts, i.e. leaves and flowers, of Calluna is observed only in the O
horizon. This “n-alkane, steroid and triterpenoids” signal is quickly lost in the
underlying Ah horizon due to (bacterial) oxidation. The other total lipid extracts
obtained are dominated by root-derived compounds. In sub-soil horizons high in
organic matter, i.e. the Ahb and Bh horizons, root-derived friedoolean and steroid
compounds dominate the total lipid signal. Organic matter poor, degraded horizons,
i.e. the E2, Bhs, Bs and B/C horizons, are dominated by C22 and C24 ω-hydroxy acids,
long-chain (>C20) n-alkanoic acids with a strong even over odd predominance and C22
and C24 n-alkanols. Steroid and root-derived triterpenoids with a friedooleanan
structure have been removed from these horizons through degradation. Based on TOC
content and lipid composition, the formation of an E1 horizon has started, but is not
yet complete. In the Ahb horizon, a contribution from buried vegetation to the total
lipid signal is still present, although degradation and an input from roots have
significantly altered the original signal. Overall, lipid data indicate that degradation
(microbial oxidation) is an important process that should be taken into account, in
addition to leaching, when describing podzolization processes in soils.
3.1 Introduction
Many soils show evidence of podzolization, a process that involves a pronounced
downwards translocation of iron, aluminum and organic matter to form characteristic
soil horizons (van Breemen & Buurman, 2002). From top to bottom, a podzol profile
consists of up to five major horizons, two of which may be absent (in italics): a litter
layer (O), a humic mineral topsoil (Ah), caused by biological mixing, a leached
eluvial layer (E), an accumulation layer of organic matter in combination with iron
and aluminum (Bh, Bhs, Bs) and thin bands with organic matter accumulation in the
subsoil (van Breemen & Buurman, 2002). Podzols form both in well-drained soils
(xero-podzols) and in soils with a shallow, fluctuating water-table (hydro-podzols).
Breakdown of litter and exudation by plants and fungi result in a large variety of
organic compounds introduced to the soil. Part of the decomposing products is soluble
in water and can be transported with percolating rainwater to deeper parts of the soil.
The remaining organic matter at the soil surface forms the humified part of the litter
layer. It can be mixed with the mineral soil to form an Ah horizon where it is further
D. F. W. Naafs, P. F. van Bergen, M. A. de Jong, A. Oonincx, J. W. de Leeuw.
European Journal of Soil Science, in press.
21
broken down by microflora, also supplying soluble organic substances. Soluble
percolating organic compounds include high molecular weight substances with
various carboxylic and phenolic groups, as well as low molecular weight (LMW)
organic acids, such as aliphatic acids and simple sugars (van Breemen & Buurman,
2002). In well-drained acid soils, iron and aluminum often move as soluble organometal chelating complexes (e.g. Schnitzer & Skinner, 1963; Jansen, 2003) or as a
positively charged mixed Al2O3-Fe2O3-SiO2-H2O sol (Farmer, 1984). Several
processes, a combination of which will play a role in all soils, may halt the movement
of dissolved organic matter and the associated metals, Al and Fe. Complexed LMW
organic acids can be decomposed by micro-organisms (Lundström et al., 2000), the
high molecular weight (HMW) organic-metal complex may become saturated with
metals and precipitate (Schnitzer & Skinner, 1963) and/or the water transport may
stop (van Breemen & Buurman, 2002). Furthermore, LMW organic compounds,
including a variety of lipids, may become associated with organic matter already
present in relatively organic-rich soil horizons, i.e. Ah and/or Bh horizons. In the
proto-imogolite theory of podzolization (Farmer, 1982), allophane and associated iron
oxides are deposited from the moving sol. Negatively charged organic sols and
solutions precipitate on the (so formed) positively charged sesquioxide-coated
surfaces of the Bs horizon to create a Bh horizon (Farmer, 1982, 1984).
Soil lipids are, by definition, organic compounds insoluble in water but soluble in
common organic solvents. They include n-alkanoic acids, n-alkanols, hydroxy acids,
ketones, steroids, terpenoids, acyl glycerols and hydrocarbons, as well as
phospholipids and lipopolysaccharides (Stevenson, 1982). These compounds originate
from both plants and animals as products of deposition, decomposition and exudation,
as well as from various other pedogenic sources, including fungi, bacteria and
mesofauna (Bull et al.., 2000a). Lipids accumulate in acid soils such as podzols
(Jambu et al.., 1985). In these, as well as in other soils, their composition is influenced
by a wide range of processes, including bioturbation, oxidation, microbial degradation
and hydrolysis. The rate of these processes is directly affected by soil pH, moisture,
microbial biomass, etc. (van Bergen et al., 1997, 1998). So far, lipid research in
podzols has mainly focussed on specific lipid fractions, e.g. n-alkanes, n-alkanols and
n-alkanoic acids, mainly in litter and A horizons (e.g. Amblès et al., 1998; Jambu et
al., 1993). To our best knowledge, no studies on the composition of total lipid extracts
throughout the whole podzol profile have been undertaken.
In this paper, the composition of total lipid extracts from 9 characteristic podzol
horizons from the Veluwe area (a natural park, located in the centre of The
Netherlands) were determined using gas chromatography (GC) and gas
chromatography-mass spectrometry (GC/MS). In addition, soil pH (H2O), ammonium
oxalate- and sodium pyrophosphate extractable aluminium and iron were determined.
Results are discussed in terms of the origin of the compounds identified and the
processes that effect their distribution in the podzol profile.
3.2 Materials and methods
3.2.1 Site description
Samples were taken from a Haplic Podzol (FAO, 1998) located at the Veluwe near
Kootwijk/Assel, The Netherlands, which has formed in wind blown sands, deposited
during the Pleistocene. Vegetation is dominated by heather (Calluna vulgaris) and
moss together with small birch trees. The profile is characterized by a litter layer (O)
underlain by several horizons typical of podzols (Fig. 3.1). recognizeable roots and
root-fragments were found only to a depth of about 52 cm (start of the Bs horizon).
22
Depth (cm)
0
6
8
18
24
O
Ah
E1
Ahb
E2
38
48
54
Bh
Bs
B/C
100 cm
Figure 3.1 Schematic representation of
the podzol profile sampled.
Characteristic horizon symbols are
indicated at the right.
3.2.2 Sampling, sample pretreatment and inorganic analyses
Samples were taken in early
September 1999 from each horizon
and from in between the Bh and Bs
horizons (referred to as the Bhs
horizon in this paper). Samples were
oven dried at 60 °C and sieved over a
2mm and a 250µm sieve to remove
roots. Soil pH (H2O) of the soil was
measured in the supernatant of a
suspension (1:2.5 sieved (<250 µm)
sample: water) using a Scott Geräte pH
meter CG 805. Sodium pyrophosphate
extractable Fe and Al were determined
by shaking 0.5 g of sieved (<250 µm)
soil overnight (16 h) with 50 ml of 0.1
M Na pyrophosphate solution.
Oxalate-extractable Fe, Al and Si were
determined by shaking (4 h in the
dark) 0.5 g of sieved (<250 µm) soil
with 25 ml 0.2 M ammonium oxalate
solution. Concentrations of Al, Fe and
Si were measured using a Perkin
Elmer Optima3000 ICP-OES.
3.2.3 Organic analyses
Total organic carbon contents (TOC%) of sieved (<250 µm) soil were measured using
a Fisons Instruments NA 1500 NCS analyzer, with a cycle time of 180 s, a source
temperature of 190 °C and an oxygen flow of ca. 30 lmin-1. Approximately 15-25 g of
the sieved (<250 µm) samples was extracted using a Soxhlet apparatus with
dichloromethane/methanol (DCM/MeOH) (9:1v/v) for 24 h. The DCM/MeOH extract
collected was taken to complete dryness in a rotary evaporator. The dry residue
obtained was dissolved in approximately 2 ml DCM/isopropanol (2:1 v/v), filtered in
a Pasteur pipette packed with defatted wool, 0.5 cm Na2SO4 and 2 cm SiO2, and dried
in a stream of N2. Free hydroxyl and carboxylic acid groups present in an aliquot were
derivatized to their corresponding trimethylsilyl (TMS) ethers and esters respectively,
by heating for 1 h at 70 oC with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA )
containing 1% trimethylchlorosilane. The derivatized aliquots were dried under N2
and dissolved in hexane. An aliquot of a standard solution containing 0.18 µg µl-1 10nonadecanone was added. The remaining part of the extract was evaporated to
dryness in air.
3.2.4 Gas chromatography (GC)
Derivitized total lipid extracts in hexane (1 µl) were analyzed using gas
chromatography (GC). GC analyses were performed using a Hewlett-Packard 6890
equipped with a CP-sil 5CB silica column (50 m x 0.32 mm, film thickness 0.12 µm).
Extracts were injected onto the column. The oven temperature was programmed from
70°C to 130°C at 20°C min-1 and from 130°C to 320°C (isothermal for 20 min at
320oC) at 4°C min-1. Compounds were detected using a flame ionisation detector
(FID) at 325°C. Helium was used as carrier gas.
23
3.2.5 Gas chromatography-mass spectrometry (GC/MS)
GC/MS analyses were performed using a Hewlett-Packard 5890 series II gas
chromatograph connected to a Fisons instruments VG platform II mass spectrometer
operating at 70 eV, scanning the range m/z 50-650 in a cycle of 0.65 seconds. The
capillary column and temperature programme were as described for the GC analyses.
Identification of the compounds was carried out using their mass spectral data and a
NIST library or by interpretation of the spectra and the GC retention times. In
addition, compound identification was based on published data (e.g. Holloway, 1982;
Killops & Frewin, 1994; van Bergen et al., 1997).
3.3 Results
3.3.1 Inorganic soil analyses
Values for pH (H2O) ranged from 4.6 to 3.5 (Table 3.1). From the O horizon (4.6) to
the Ah horizon (3.7), the pH (H2O) decreased, only to rise again to 4.6 in the E2
horizon. Active iron (Feox) was highest in the Bs horizon and lowest in the E2
horizon. The amount of organically complexed Al (Alpy) was highest in the Bh
horizon and, like the organically-complexed Fe, was low in the E2 horizon.
Table 3.1 Inorganic soil analyses
Horizon
O
Ah
E1
4.6
3.7
3.9
pH (H2O)
0.11 0.31 0.02
Alox%
0.12 0.14 0.04
Alpy%
0.07 0.23 0.02
Feox%
0.07 0.17 0.03
Fepy%
Ahb
3.8
0.23
0.10
0.18
0.07
E2
4.6
0.02
0.02
0.01
0.02
Bh
3.5
1.81
0.82
0.87
0.42
Bhs
3.5
0.33
0.28
0.61
0.52
Bs
4.0
0.84
0.39
1.54
0.75
B/C
4.1
0.41
0.15
0.07
0.03
3.3.2 Total organic carbon (TOC) and quantification of total lipid extracts
The largest TOC contents (>5%) were found in the O, Ah, Ahb and Bh horizons,
while both E horizons together, with the Bs and B/C horizons, contained less than 2%
TOC (Table 3.2). TOC contents of residues after solvent extraction showed a similar
trend to those of the unextracted samples. The yields of total lipid extracts (mg g-1)
soil ranged from 0.02 to 2.8, the highest amounts again found in the humic horizons.
The Ah horizon was making the largest contribution of lipids to the TOC on a mass
basis.
Table 3.2 Organic soil analyses
O
Ah E1
Ahb E2
Bh Bhs
Horizon
TOC (wt%)
10.1 7.8 2.0
5.3 0.6 6.6 1.7
TOC Sox. Res. (wt%)
9.6 7.3 1.6
4.9 0.5 5.7 1.2
TLE (mg.g-1 soil)
2.2 2.8 0.3
1.1 0.1 1.6 0.2
TLE (mg.g-1 TOC)
21.9 35.7 13.0 20.8 22.3 24.2 12.9
24
Bs B/C
1.2 0.4
1.1 0.4
0.1 0.1
3.9 6.7
3.3.3 The composition of total lipid extracts from all 9 characteristic horizons
The extract from the O-horizon was dominated by a series of long-chain (>C25) oddover-even dominated n-alkanes (Fig. 3.2a), the most abundant being C31 and C33. In
addition, steroids and triterpenoids (Table 3.3) were detected including taraxerol,
camposterol, stigmasterol, β-sitosterol, C29 ketosteroid, α-amyrin, lupeol, friedoolean3-one, an unidentified triterpenoid and a C30 triterpenyl acid. Other compounds,
present in relatively lower concentrations, were even-over-odd predominated C18-C26
n-alkanols, n-alkanoic acids, ranging from C14-C26, C15 iso and anteiso acids, C18
isoprenoid methylketone and short-chain (<C25) n-alkanes without an odd-over-even
predominance.
C31 and C33 n-alkanes, together with their co-eluting C29 and C31 methylketones,
were also relatively abundant in the extract of the Ah horizon (Fig. 3.2b). In addition,
a C29 ketosteroid together with triterpenoids including taraxerone, friedoolean-3-one
and an unknown steroid dominated the last part of the chromatogram. Even-over-odd
dominated C12-C32 n-alkanoic acids and C23-C28 odd-over-even dominated
methylketones were found in relatively large concentrations. Apart from saturated
linear fatty acids, C15 iso and anteiso acids and one C18 n-alkenoic acid were found. βsitosterol was present only in relatively low concentrations, together with even C20C26 n-alkanols, the C18 isoprenoid methylketone, C22 and C24 ω-hydroxy acids, and
short-chain (<C25) n-alkanes without an odd-over-even predominance.
Table 3.3 Characteristic mass fragments and IUPAC names of steroidal and known
and unknown triterpenoid compounds (analyzed as their TMS derivatives)
Compound (in order of elution)
Characteristic fragment ions (m/z) [M]+
69, 95, 205, 218, 231, 243, 257, 274,
410
Friedoolean-2-ene1 (T1)
Unidentified triterpenoid (T2)
24-Methylcholest-5-en-3β-ol1, 2
(Camposterol) (S-C28:1)
Friedoolean-14-en-3-one (T3)
(Taraxerone)1
Friedoolean-14-en-3β-ol (T4)
(Taraxerol)1
24-Ethylcholest-5, 22-dien-3β-ol1, 2, 3
(Stigmasterol) (S-C29:2)
24-Ethylcholest-5-en-3β-ol1, 2, 3
(β-sitosterol) (S-C29:1)
24-Ethylcholestan-3β-ol (S-C29)1
Ursan-12-en-3β-ol2 (α-Amyrin) (T5)
Lupeol2, 3 (T6)
Steroid1 (S1)
Friedooleanan-3-one1 (T8)
Triterpenoid1 (T9)
287, 395
73, 190, 204, 218, 269, 359, 393, 483
75, 129, 213, 255, 261, 343, 367, 382,
457
69, 81, 95, 109, 189, 191, 203, 205,
273, 409
121, 135, 189, 204, 269, 284, 359, 374,
393, 408, 483
83, 129, 255, 351, 379, 394, 469
484
129, 255, 275, 357, 381, 396, 471
486
107, 215, 257, 305, 359, 383, 398, 473
73, 95, 189, 190, 203, 218, 279, 393,
408, 483
73, 190, 203, 218, 369, 393, 408, 483
95, 109,187, 205, 259, 274, 286, 393
69, 81, 95, 109, 123, 273, 302, 341,
379, 411
69, 81, 95, 109, 121, 137, 205, 218,
313, 327, 341, 409, 424, 481
69, 81, 95, 109, 123, 147, 163, 177,
207, 231, 257, 275, 317, 395
73, 133, 189, 203, 279, 320, 483, 585
488
498
498
472
424
498
498
408
426
496
414
24-Ethyl-5-cholestan-3-one (KS-C29)1
(C29-ketosteroid)
600
C30 Triterpenyl acid3 (Ta1)
1
2
3
(van Smeerdijk and Boon, 1987), (Killops and Frewin, 1994), (van Bergen et al.,
1997), letters (Tx, Sx) used in Fig. 3.2 shown in parenthesis.
25
Fig. 3.2 Gas chromatograms of total lipid extracts (TLEs) obtained from: O-horizon
(a), Ah horizon (b), E1 horizon (c), Ahb horizon (d), E2 horizon (e), Bh horizon (f),
Bhs horizon (g), Bs horizon (h) and B/C horizon (i). Key: : n-alkanols, : n-alkanoic
acids, : n-alkanes, : ω-hydroxy acids, : ketones, su : saccharide, I-C18 mk:
isoprenoid methylketon, Ph: phthalate. Cx refers to the total number of carbon atoms.
S-Cx refers to steroids and KS-Cx refers to keto-steroids with x referring to the total
number of carbon atoms. Number after colon refers to the total number of double
bonds. Tx and Tax refers to triterpenoids and triterpenoic acids respectively, with x
referring to their number as assigned in the text and Table 3.
26
Figure 3.2 d, e, f.
The only difference between the extract from the E1 horizon (Fig. 3.2c) and that
from the Ah horizon (Fig. 3.2b) was the increase in relative concentration of the nalkanoic acids. In contrast, these acids were much less important in the Ahb extract
(Fig. 3.2d), which extract was characterized both by β-sitosterol and the C29
ketosteroid. Other important compounds were friedoolean-2-ene, T2, friedoolean-3one, T9, stigmasterol, C22 and C24 n-alkanols, and the co-eluting C31 n-alkane and C29
methylketone.
In the E2 horizon, the relative contribution of steroids and triterpenoids was
smaller again, whereas the even-over-odd dominated n-alkanoic acids, n-alkanols and
C22 and C24 ω-hydroxy acids were more abundant (Fig. 3.2e). Two short-chain, i.e. C8
and C9, ω-hydroxy acids were identified together with phytol, C23-C31 methylketones
and C12 and C14 n-alkanoic acids.
27
Fig. 3.2 g, h, i.
β-sitosterol and a C29 ketosteroid, together with a number of triterpenoids including
friedoolean-2-ene, taraxerone, taraxerol and friedoolean-3-one, characterized the
extract obtained from the Bh horizon (Fig. 3.2f), similar to the Ahb horizon extract
(Fig. 3.2d). Other relatively abundant compounds included C22 and C24 n-alkanols and
n-alkanoic acids. In addition to a series of odd-over-even dominated ketones (C23C31), n-alkanes without a clear odd-over-even dominance (C20-C23) were identified.
C22 and C24 ω-hydroxy acids were found to dominate the extract obtained from the
Bhs horizon in addition to an even-over-odd dominated series of n-alkanoic acids
28
(C10-C28) and n-alkanols (C14-C28; Fig. 3.2g). Compounds present in relatively smaller
concentrations included n-alkanes (C18-C23), β-sitosterol, taraxerone, friedoolean-3one, the C29 ketosteroid, as well as ketones (C23-C29) and two saccharides.
The Bs-horizon total lipid extract was characterized by a series of even-over-odd
dominated C12-C28 n-alkanoic acids maximizing at C18, C22 and C24 (Fig. 3.2h). In
addition, a series of even-over-odd dominated n-alkanols ranging from C14 to C28 and
maximizing at C20, C22 and C2,4 was found. Other compounds detected were βsitosterol, C22 and C24 ω-hydroxy acids, C27 and C29 ketones, a saccharide, taraxerone
and friedoolean-3-one.
Like the sample obtained from the Bhs horizon (Fig. 3.2g), long-chain (>C20) nalkanoic acids together with C22 and C24 ω-hydroxy acids characterized the B/C
horizon extract (Fig. 3.2i). Although there was a slight increase in the relative
concentration of C20 n-alkane and the long-chain n-alkanoic acids, the B/C extract
resembled to a great extent the extract obtained from the overlying Bhs horizon (Fig.
3.2h).
3.4 Discussion
3.4.1 Inorganic and bulk organic soil analyses
The largest amounts of aluminium and iron complexed with organic matter, i.e. Alpy
and Fepy are found in the B horizons (Bh, Bs; Table 3.1). According to these data,
soluble organic matter transported down the profile is immobilized in particular by
aluminium in the Bh horizon (Lundström et al ., 1995; Wesselink et al., 1996). The
Alox and Feox data, in combination with Alpy and Fepy data, show that the Bs horizon is
mainly formed by Fe oxides (Table 3.1).
Acid soils are known to enhance the preservation of lipids (Jambu et al., 1985; van
Bergen et al., 1998; Bull et al., 2000a). All concentrations of extractable lipids, with
the exception of the Ah sample, where it is 35.7 mg g-1 TOC (Table 3.2), lie within
the range of 2-25 mg g-1 TOC normally found for soils (Stevenson, 1982). The low
soil pH therefore seems to have no great influence on the preservation of lipids in the
soil profile studied.
The change of TOC% with depth (Fig. 3.3) clearly indicates the presence of
organic matter poor layers (E1 and E2), an accumulation of organic matter (Ah, Bh),
and a litter layer (O). The Bhs, Bs and B/C samples contain minor amounts of organic
matter, indicating that its accumulation occurs mainly above the hard cemented Bs
horizon (de Coninck, 1980). In addition to these characteristic horizons (van Breemen
& Buurman 2002), accumulated organic matter was found in the buried A horizon
(Ahb). The newly formed E1 horizon on top of this Ahb horizon (Fig. 3.1) still
contains 2% TOC (Table 3.1 & Fig. 3.3) suggesting that the formation of this horizon
has started, but is not yet complete.
29
Depth (cm)
TOC%
0 2 4 6 8 10
0
O
Ah
10
E1
20
Ahb
30
E2
40
Bh
Bh/s
50
Bs
B/C
60
1
0
3
2
TLE mg.g-1 soil
Figure 3.3 Total organic carbon
-1
content (TOC%) and TLE (mg.g dry
soil) vs depth. TOC% is represented by
the solid line, TLE by the dashed line.
Characteristic horizons are indicated
between dashed lines, horizon codes
are found at the right.
The absolute amounts of extracted lipids (mg
g-1 soil; Fig. 3.3), show a good correlation with
the TOC contents (Fig. 3.4). When the Ah
horizon is excluded, the correlation is even
better (r2 = 0.97) showing that lipids
accumulate in organic matter rich horizons.
Comparison of the contribution of lipids to
TOC with the TOC content reveals that lipids
do not accumulate relative to other soil organic
matter in these horizons. In other words, SOM
compounds other than lipids have accumulated
even more than lipids in podzol B horizons
(Schmidt et al., 2000). The relatively great
contribution of lipids to the total amount of
organic carbon in the E2 horizon (Table 3.2)
could be caused by the relatively low
leachability of hydrophobic compounds, such
as lipids, compared with more hydrophilic
organic compounds. Alternatively, it could be
due to the very low TOC content which would
exaggerate any errors in this ratio. The
difference between the theoretical yield
calculated using the difference in TOC%
between the unextracted and solvent extracted
samples (Table 3.2), and the absolute weight of
the total lipid extracts, is caused by losses
during the work-up procedure of the extract.
The large contribution of free lipids in the Ah
horizon parallels the accumulation of aliphatic
compounds in A horizons of acidic soils
observed previously (Jambu et al., 1993) due to
the hydrolysis of wax esters (Jambu et al.,
1993).
TLE (mg.g-1 dry soil)
3.4.2 Molecular composition of total lipid extracts (TLEs) from characteristic
horizons
Composition of the TLE from the O horizon. Dominant long-chain (>C27) alkanes
with a strong odd-over-even predominance (Fig. 3.2a) are traditionally ascribed to
epicuticular waxes and protective layers on vascular plants and commonly observed
in lipid extracts from vegetation and soils (Jambu et al., 1991; Amblès et al., 1989).
3
2.5
2
1.5
1
0.5
r = 0.94
0
0
2
4
6
TOC%
-1
Fig. 3.4 TLE (mg.g dry soil) vs TOC%
30
8
10
12
ug.g-1 dry soil
Moreover, the relatively high amounts of C31 and C33 alkanes (Figs 3.6c, 3.7c),
with smaller contributions of C27 and C29, are characteristic of heather (Calluna)
flowers and leaves (Nierop et al., 2001). A heather-derived input is also reflected in
the C22 and C24 n-alkanol distribution, characteristic for Calluna stem wood and roots
(Nierop, 1998; Nierop et al., 2001) as well as Ericaceae rootlets (van Smeerdijk &
Boon, 1987).
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
ug.g-1 TOC
1
0.8
a)
O
C29:1
C29 keto
Ah
E1
Ahb E2
Bh
Bhs Bs
B/C
Ah
E1
Ahb E2
Bh
Bhs Bs
B/C
b)
0.6
0.4
0.2
0
O
Figure 3.5 Absolute concentration (ug.g-1 dry soil) (a), and relative concentration
(ug.g-1 TOC) of b-sitosterol and C29 ketosteroid (b), per horizon.
The n-alkanoic acids identified can have various natural origins such as plant,
fungal or bacterial origins (Amblès et al., 1994). Straight-chain compounds of fungal
origin are similar to those of plant origin, but they only range from C10 to C24, often
characterized by the presence of unsaturated C16 and C18 acids (Weete, 1974; Ruess et
al., 2002). Furthermore, long-chain (>C20) n-alkanoic acids in this top-soil layer may
originate from the hydrolysis of wax esters (van Bergen et al., 1998) or the oxidation
of a variety of other compounds such as n-alkanes and n-alkanols (Mouçawi et al.,
1981; Amblès et al., 1994a, b), processes which are enhanced by a low soil pH (van
Bergen et al., 1998a) and the presence of hydrous iron oxides (Mouçawi et al., 1981).
The strong even-over-odd dominance and chain-length distribution observed (Fig.
3.2a) imply that odd-chain compounds such as n-alkanes are unlikely sources.Based
on the same criteria, and the dominance of both C22 and C24 acids, a contribution from
the oxidation of n-alkanols is more likely.
The n-alkanes ranging from C20 to C27, with a carbon preference index close to
unity, are more probably of microbial origin (Amblès et al., 1989), although they have
also been identified in association with Ericaeae rootlets (van Smeerdijk & Boon,
1987). A microbial input is also obvious from the iso- and anteiso- C15 fatty acids
(Boon et al., 1977), and possibly by the detection of C18 alkenoic acid (Parlanti et al.,
1994).
31
The steroids and triterpenoids identified have frequently been detected in soil and
leaf or litter extracts (Killops & Frewin, 1994; van Bergen et al., 1997; Bull et al.,
1998). However, it is important to note that lupeol, taraxerol, α-amyrin and the
triterpenoic acid are more associated with aerial vegetation (Killops & Frewin, 1994).
On the other hand, all steroid compounds identified, together with friedoolean-3-one
and T9, have been recognized as significant constituents of heather roots (van
Smeerdijk & Boon, 1987; Nierop et al., 2001), although heather leaves and flowers do
not contain these compounds in significant amounts (Nierop et al., 2001). In addition
to their vascular plant origin, C29 steroids are known to be derived from fungi (Weete,
1976; Grandmougin-Ferjani et al., 1999). Keto-steroids are oxidation products of
sterols (Goad, 1991) and have been identified in association with heather rootlets (van
Smeerdijk & Boon, 1987). The C18 isoprenoid methylketone is a degradation product
of phytol, which has been found before in total lipid extracts from aerial vegetation
(van Bergen et al., 1997; Bull et al., 2000a).
Composition of the TLE from the Ah horizon. Both the relative (Figs 3.2b, 3.5b)
and absolute (Fig. 3.5a) amounts of β-sitosterol decrease from the litter to the Ah
horizon. A decrease in sitosterol, as well as other steroids and terpenoids associated
with aerial parts of vegetation (Killops & Frewin, 1994), was seen when lipids from
forest soil litter layers or leaf extracts were compared with those from the underlying
A horizon (Jambu et al., 1993; Amblès et al., 1994a, b; van Bergen et al., 1997; Bull
et al., 2000a). It is suggested that these compounds are easily mineralized in the soil
(van Bergen et al., 1997) or that their decrease in soluble lipid fractions can be caused
by their condensation into more stable, insoluble moieties (Gobé et al., 2000). In
addition, they may be chemically altered to form modified steroids and triterpenoids
(van Bergen et al., 1997).
The significant increase in concentration of the C29 ketosteroid (Fig. 3.5) and
taraxerone from the O to the Ah horizon probably results from oxidation (van Bergen
et al., 1997). In addition, taraxerone (T3), friedoolean-3-one (T8) and T9 triterpenoids
have also been found in heather rootlets (van Smeerdijk & Boon, 1987; Nierop et al.,
2001) reflecting a direct input. In general, more unsaturated compounds are more
susceptible to (microbial) degradation. The total decrease of stigmasterol (C29:2)
compared with that of β-sitosterol (C29:1) is therefore related to the degree of
unsaturation. In addition, a source of β-sitosterol has been found in ester-linked lipids
obtained from soil samples taken from a depth of 15-50 cm (Chapter 6).
Another group of lipids that decreases rapidly from the O to the Ah horizon are
the n-alkanes (Figs 3.6c, 3.7c), as observed in acid soils (Marseille et al., 1999). This
decrease in n-alkane concentration, is accompanied by an increase in the
concentration of methylketones (Figs 3.2b, 3.6d & 3.7d), which are not known as
primary plant products. Considering that n-alkanes are a likely substrate for in situ
microbial β-oxidation based on similarities found in the distribution of ketones and nalkanes (Amblès et al., 1993), and that such similarities are found in the podzol
profile (Fig. 3.2a), it is assumed that methylketones in the Ah horizon are derived
from such oxidation. The relatively minor contribution of methylketones to the total
amount of lipids in the litter layer (Fig. 3.6d) is expected because methylketones are
usually only detected in such layers in trace amounts (Amblès et al., 1993).
Furthermore, because of their biodegradation, the amount of methylketones
determined in soil is much lower than the amounts produced from n-alkanes (Amblès
et al., 1993).
32
Bhs
Bs
1
C24
Bh
Bhs
Bs
B/C
C29
C31
ketones
0
C22
C24
C16
C24
C16
C24
C
C2422
C26
C28
C16 C
22
B/C
Bhs
Bs
B/C
Bhs
Bs
B/C
Bs
B/C
Bs
B/C
C24
c)
2
C22
C24
C
C2422
C
C2422
C22
C24
C22
C24
Ah
E1
Ahb
E2
Bh
1,5
C20
C20
1
0,5
0
O
0,4
d)
0,2
Ah
E1
0,1
Ahb
E2
Bh
E2
Bh
C31
0
C26
C
C24 22
0,005
C22
Bs
0
O
Ah
E1
Ahb
E2
Bh
Bhs
B/C
ω-hydroxy acids
e)
0,01
Bhs
Bs
B/C
Figure 3.6 Absolute concentration (ug.g-1
dry soil) of C16-C33 lipid compounds per
horizon: (a) n-alkanoic acids, (b) n-alkanols,
(c) n-alkanes, (d) ketones and (e) w-hydroxy
acids.
1
0,8
O
Ah
E1
Ahb
E1
Ahb
e)
Bhs
C22
Bh
C24
E2
0,6
0,4
0,2
C26
Ahb
C22
C24
E1
C
C2422
Ah
C24
O
ω-hydroxy acids
0,015
O
2,5
0,3
0,01
Bs
ketones
0,02
Bhs
C29
C31
E2
C29
d)
Ahb
Bh
29
C20
C33
C29
C31
E1
E2
C27C
n-alkanes
0
Ah
Ahb
0
C33
B/C
C31
Bs
0,2
C29
Bhs
0,4
C27
Bh
0,1
0,05
E1
0,6
C22
C24
C22
C24
E2
Ah
C31
C33
Ahb
Relative concentration (ug.g-1 TOC)
E1
O
b)
0,8
C
C2422
C
C24 22
Ah
C16
0
C22
0,15
0,03
0,2
B/C
c)
O
0,4
C22
Bh
C31
C33
O
C24
C16
C24
C22
C24
C26
E2
C24
C24
n-alkanes
0,2
C
C2422
0,03
0,25
Ahb
b)
0,04
0,02
0,01
0
Ah
n-alkanols
Absolute concentration (ug.g-1 dry soil)
O
C24
a)
0,6
n-alkanols
0,06
0,05
C24
C16
0,04
0,03
0,02
0,01
0
0,8
C24
n-alkanoic acids
a)
n-alkanoic acids
0,06
0,05
0
O
Ah
E2
Bh
Bhs
Figure 3.7 Relative concentration
(ug.g-1 TOC) of C16-C33 lipid
compounds per horizon: (a) n-alkanoic
acids, (b) n-alkanols, (c) n-alkanes, (d)
ketones and (e) w-hydroxy acids.
The increase in absolute (Fig. 3.6a) and relative (Fig. 3.7a) concentrations of longchain (>C20) n-alkanoic acids (cf. Amblès et al., 1994b), together with the increase in
the total amount of lipids (Fig. 3.3), indicates that these compounds are produced in
the Ah horizon. Hydrolysis of wax esters (van Bergen et al., 1998) and an input from
roots (Bull et al., 2000b) are the most likely processes involved in this production.
The presence of hexadecanoic acid and C18 n-alkenoic acid could be related to an
input from mosses (Nierop et al., 2001), but has also been observed in root- and
rhizosphere-derived total lipid extracts (Bull et al., 2000a, b). Moreover, (<C20) acids,
including C14, C16, C18:1 and C18 fatty acids, were found to be the dominant acids
leached from the litter and Ah horizon during an experimental study of podzol soil
lipids (Amblès et al., 1998).
33
The absolute (Fig. 3.6b) and relative (Fig. 3.7b) increase in the concentration of nalkanols is most likely to be the result of both an input from heather roots, considering
their relatively large abundance in heather root pyrolysates (van Smeerdijk & Boon,
1987; Nierop, 1998; Nierop et al., 2001) and from the hydrolysis of wax esters
(Jambu et al., 1993). The latter process is enhanced by the low soil pH (Wolfe et al.,
1989). The detection of C22 and C24 ω-hydroxy acids (Figs 3.6e and 3.7e) in this
horizon reveals an input derived possibly both from aerial vegetation as well as from
roots (Bull et al., 2000b).
Composition of the TLE from the E1 horizon. As mentioned, the composition of
the total lipid extract from the E1 horizon (Fig. 3.2c) resembles to a great extent that
from the overlying Ah horizon (Fig. 3.2b). Taking into account the relatively high
TOC content (2%) for a leached horizon (Fig. 3.3), this strongly suggests that the
process of formation of this E1 horizon is not yet complete, although both absolute
and relative amounts of lipids are very small (Figs 3.6, 3.7). Normalized to the TOC%
there is, apart from the decrease in n-alkanoic acids and n-alkanols, a strong decrease
in the concentration of ketones (Fig. 3.7d) caused by a relative rapid biodegradation
as intermediates in the n-alkane degradation pathway (Amblès et al., 1993).
Composition of the TLE from the Ahb horizon. The Ahb extract again shows a
signal dominated by steroids and triterpenoids (Fig. 3.2d), with abundant
contributions from β-sitosterol (see also Fig. 3.5), C29 ketosteroid, and friedoolean-2ene (T1), T2, friedoolean-3-one (T8) and T9 triterpenoids. All these compounds have
been identified in heather rootlets (van Smeerdijk & Boon, 1987). Relative to the
TOC%, β-sitosterol (Fig. 3.5b), n-alkanols (Fig. 3.7b), and ketones (Fig. 3.7d) show a
significant increase. These compounds are enriched in the buried A-horizon either as
remnants of the buried vegetation, or as the result of an input by roots (Nierop et al.,
2001). A contribution from above-ground litter lipids through transport to Ah
horizons in sandy soils, on the other hand, has been found to be negligible. Possible
evidence for a contribution from previous vegetation could be the increase in both
absolute (Fig. 3.6d) and relative (Fig. 3.7d) methylketone concentrations most
probably derived from the oxidation of n-alkanes (Amblès et al., 1993; van Bergen et
al., 1997).
Composition of the TLE from the E2 horizon. C22 and C24 ω-hydroxy acids, in
combination with long-chain (>C20) acids and C22 and C24 n-alkanols, are the lipids
that remain in the E2 horizon. In addition to these root-derived compounds (Nierop,
1998; Chapter 2), C8 and C9 ω-hydroxy acids were detected, probably as a result of
the oxidation (Regert et al., 1998) of unsaturated suberin building blocks (Nierop et
al., 2003). It should be noted that the distributions of both n-alkanoic acids and ωhydroxy acids resemble those observed in samples of rhizosphere and mineral
horizons (Bull et al., 2000a, 2000b).
Methylketones also remain in this horizon as oxidation products of alkanes
(Amblès et al., 1993). Due to this intensive leaching or degradation, absolute amounts
of all compounds identified in this horizon are very small (Fig. 3.6). Normalized to
TOC%, however, it is shown that the decrease in the contribution of lipids to the
TOC% is mainly caused by a decrease in steroids (Fig. 3.5b), triterpenoids and
ketones (Fig. 3.7d). Considering the labile nature of these compounds (van Bergen et
al., 1997) and their hydrophobicity, (microbial) degradation of these compounds
seems the most likely explanation for this decrease. The contributions of n-alkanoic
acids (Fig. 3.7a), n-alkanols (Fig. 3.7b) and ω-hydroxy acids (Fig. 3.7e) on the
contrary remain relatively similar, or even increase. These compounds therefore
reflect a root-derived input that is hard to degrade (and/or leach). No evidence of
accumulation of lipids derived from the soil surface, such as long-chain (<C20) nalkanes, has been observed in the E2 horizon. Thus, the E2 signal could be considered
to be derived of an input by roots.
34
Composition of the TLE from the Bh horizon. Just as observed for the O and Ahb
horizon, the organic matter rich Bh horizon is characterized by β-sitosterol and C29
ketosteroid, a series of triterpenoids, and C22 and C24 n-alkanols (Fig. 3.2f).
Comparison with the large TOC content (Fig. 3.3) and both absolute and relative
amounts of lipid compounds identified, clearly indicates that organic matter, including
lipids, accumulate in this horizon. Podzol B-horizons may contain more root-derived
than illuviated organic matter (van Breemen & Buurman, 2002).
The contribution of litter-derived organic matter to the sub-soil in sandy soils was
found to be negligible (Nierop et al., 2001). Moreover, roots were found to develop
preferentially in podzol Bh horizons (de Coninck, 1980). Thus, the significant
increase in the contribution of steroids (Fig. 3.5), triterpenoids and C22 and C24 nalkanols (Figs 3.6b, 3.7b), all known to be derived from heather roots (van Smeerdijk
& Boon, 1987; Nierop & Buurman, 1999; Nierop et al., 2001), can be interpreted as
resulting from the accumulation of root-derived lipids produced in situ. In addition to
this root-derived signal, a bacterial contribution is again reflected in the presence of
short-chain alkanes and methylketones. It should be noted that, similar to the
observations by Nierop & Buurman, (1999) with respect to the insoluble organic
matter fraction, no evidence was found for a contribution than lipid compounds
transported from the top of the profile.
Composition of the TLE from the transition from the Bh to the Bs horizon. In the
sample from the transition layer between the Bh and Bs horizons (Bhs; Fig. 3.2g), the
total lipid signal again changes dramatically into a signal dominated by ω-hydroxy
acids (Figs 3.6e, 3.7e) and n-alkanoic acids, as observed for the E2 horizon (Fig.
3.2e). The sharp decrease in TOC content (Fig. 3.3) indicates that no accumulation of
organic matter took place in this horizon, again strongly suggesting that the signal
observed is characteristic for a degraded root-derived input. The microbial signal in
this horizon comprises short-chain alkanes and methylketones. It has been suggested
that complexed OM is degraded in this horizon to yield Al and Fe which then
precipitates in the underlying Bs horizon (Lundström et al., 1995). We suggest that
this degradation may also effect the biopolyester suberin, resulting in the release of
relatively large amounts of long-chain (>C20) ω-hydroxy acids (Figs 3.5e, 3.6e)
characteristic of an input by roots to the lipid signal in soils (Bull et al., 2000b;
Chapter 2).
Composition of the TLEs from the Bs and B/C horizons. Both the TOC contents
(Fig. 3.3) and absolute amounts of lipids (Figs 3.3, 3.6) are very small in the Bs
horizon (Fig. 3.2h). The signal observed most probably reflects the very small
remnant of lipids or a minor contribution from lipids that penetrate into this hard layer
from the overlying Bhs horizon (Fig. 3.2g). Again, most root-derived steroids and
triterpenoids have been degraded.
The total lipid signal from the B/C horizon (Fig. 3.2i) again resembles that of the
other horizons with a small TOC content, i.e. the E2, Bhs and Bs horizons (Fig. 3.3).
The lipid signal consists of compounds derived from degraded remnants of old roots
that were present in this horizon before it was cut off from the rest of the profile by
the formation of the hard, cemented Bs horizon (de Coninck, 1980).
Implications of lipid data for the podzolization process. The lipids in the Bh
horizon are mainly derived from a fresh input by roots that accumulate in this horizon.
The limited penetration of surface-vegetation derived lipids into the profile indicates
that the influence of vertical transport of lipids as compared to (bacterial) degradation
is negligible. The detection of lipid degradation products such as ketones, and
bacterial markers such as short-chain alkanes and fatty acids throughout the profile,
further indicates that (bacterial) degradation of lipids is an important process in this
podzol profile. Considering that both steroids and triterpenoids have been shown to
degrade rapidly within soils (van Bergen et al., 1997), this is a likely explanation for
35
their small concentration in podzol E horizons. After degradation of most of these
steroids and triterpenoids, a signal derived from bacteria and root-derived long-chain
fatty acids and ω-hydroxy acids remains in the E horizon. These latter compounds
reflect in situ lipid input produced by the degradation of remnants of root suberins and
remain after degradation of the rest of the root-derived lipids.
3.5 Conclusions
The composition of total lipid extracts from 9 characteristic horizons in a podzol from
the Veluwe area (The Netherlands) has been analysed by gas chromatography (GC)
and gas chromatography-mass spectrometry (GC/MS). In addition, soil pH (H2O),
ammonium oxalate- and sodium pyrophosphate extractable aluminium and iron have
been determined. From these analyses it is concluded that:
1) Lipids accumulate in horizons with high organic matter contents, but there is no
increase relative to other soil organic matter.
2) An n-alkane, steroid and triterpenoid signal related to the aerial parts, i.e. leaves
and flowers, of Calluna is observed only in the O horizon. This signal is quickly lost
in the underlying Ah horizon due to (bacterial) oxidation.
3) All total lipid extracts, except that from the O-horizon, are dominated by rootderived lipids.
4) In sub-soil horizons with high organic matter contents, i.e. the Ahb and Bh
horizons, root-derived friedoolean and steroid compounds dominate the total lipid
signal.
5) C22 and C24 ω-hydroxy acids, long-chain (>C20) n-alkanoic acids with a strong
even-over-odd dominance and C22 and C24 n-alkanols, characteristic of an input from
roots, dominate the horizons with low contents of organic matter, i.e. the E2, Bhs, Bs
and B/C horizons. Steroid- and root-derived triterpenoids with a friedoolean structure,
once present in these horizons, have been degraded.
6) Based on TOC content and molecular lipid composition, the formation of an E1
horizon has started, but is not yet complete.
7) In the Ahb horizon, a contribution from buried vegetation to the total lipid signal is
still present, although degradation and an input by roots have significantly altered the
original signal.
8) Overall, lipid data indicate that degradation of lipids (mainly oxidation) is an
important process that should be taken into account, in addition to leaching, when
describing the so called “podzolization process” in soils.
Acknowledgements. This study was supported by the Institute for Paleoenvironment
and Paleoclimate Utrecht (IPPU). M. van Alphen and T. Zalm are thanked for their
help during GC and GC/MS analyses. Drs. J. Kool is thanked for supplying additional
data and graphs. Two reviewers are thanked for their useful comments and
suggestions. Both the associate editor Dr. L. Peterson and the editor-in-chief Dr. P.
Loveland of the European Journal of Soil Science are sincerely thanked for their help
with the final editing and formatting, as well as useful comments and suggestions on
the contents of the paper.
36